Mechanical ventilation (MV) is the most frequently used life-support modality in intensive care units (ICUs). To reduce anxiety, stress and promote tolerance of MV, nurses frequently administer a variety of sedative medications to their patients based on subjective observations. While sedation is indicated at times, it can have serious side effects, including delayed weaning, mental status changes, prolonged MV, and extended ICU stay. Alternative strategies are needed that are safe yet alleviate the distressful symptoms experienced by ICU patients. Patient-controlled sedation (PCS) is a strategy that warrants testing in the ICU with ventilated patients. Patient-controlled analgesia (PCA) has been widely used in care of post-operative patients since the 1980's and has demonstrated superior pain management and patient satisfaction over nurse-delivered analgesia. While PCS holds great promise with ventilated ICU patients, it has not been evaluated for longer than 24 hours, thus it is not known if PCS is safe and feasible for ventilated patients to use for symptom self- management during a significant portion of ventilatory support. This pilot study will: 1) Establish the feasibility of subject recruitment and protocol adherence in a randomized pilot trial that compares PCS using dexmedetomidine to usual sedation practice in mechanically ventilated patients, 2) Evaluate and refine the study protocol to: a) improve completion rates for subject self-assessments such as anxiety and delirium scales, b) reliably determine when subjects are unable to self-manage their sedation due to weakness, confusion or severe illness; c) create rules for suspending PCS therapy in a future RCT, and 3) Estimate effect sizes on key variables to inform the design of a future randomized trial to compare: a) sedative exposure, anxiety levels, and delirium occurrence between groups, b) adverse event rates and protocol violations related to the drug, pump or both; and c) duration of mechanical ventilation and ICU stay. A sample of 40 mechanically ventilated adult patients will be recruited from two ICUs in the urban Midwest. Subjects will be allocated to PCS or usual care whereby sedative therapy is administered by patient-care nurses. Daily assessments include visual analog scale-anxiety, alertness and delirium presence (CAM-ICU). Long-term objectives of this research are to positively impact important clinical outcomes and quality of life for the millions of ICU patients who experience MV each year by directly encouraging patient self-management of common distressful symptoms associated with ventilatory support. This project addresses NINR's area of opportunity for testing strategies to improve management of symptoms over disease trajectories (e.g., critical illness requiring MV). The proposed study takes the innovative approach of involving patients directly with symptom management. This study will obtain important preliminary data on the safety and feasibility of PCS to inform a future randomized trial directly comparing PCS with nurse-administered sedative medications on salient patient outcomes.